We’ve had so many threads on this, and it seems, to me, like it comes down to even directivity (which would certainly imply matched pairs as well once in stereo) and an individual preference for wide or narrow directivity, which has inevitable trade offs (precision, perceived ‘size’, seating flexibility). I haven’t seen any evidence that something else is at play (in the loudspeaker, that is).
As always, I could be wrong.
Yeah, this is where I'm at. I've contributed dozens of quasi anechoic speaker measurements and been reading about speakers here for four or five years, read a bunch of papers, and I still haven't seen anything convincing to indicate that the qualities (or flaws) in the speaker's soundstage comes from anything beyond directivity behavior, frequency response, and potential resonances (which are usually noticeable from the previous too), all assuming a reasonable pair match. Neither in my own listening experience nor in what I've read from others.
In fact, my issue is that I generally think people don't pay
close enough attention to directivity beyond a quick glance at the beamwidth or shape of the polar response and DI/ERDI curves in the spinorama.
I think there is potential for much more insight we can learnfrom the SPL response at individual angles and averaged responses over small windows.
There's still a lot we don't know about vertical directivity. There's still a lot to assess about the shape of the off-axis curves, such as whether they have a smoothly decreasing slope (like Neumann's and Genelecs), whether they have constant directivity for a portion of the response (like many JBL horns/), how smooth the individual off-axis curves are, what directivity is like below 1000Hz (perhaps contributing to a sense of envelopment), etc.
Then there's the matter of "wide" vs "narrow" traditional designs, let alone more esoteric ones. Narrower speakers a la Neumann, Genelec, and kef almost always have prettier off-axis curves, but then the off axis response is less similar in slope to the on-axis response than it would be on a wider speaker
Constant directivity speakers like the D&D8C have narrower directivity but maintain that relatively flat off axis slope. This seems to be something I subjectively enjoy.
(All of this is why I don't like contour plots, because I feel they obscure a lot of this information for a visual that's slightly easier to digest.)
This is all complicated by room interaction, but it's why I personally think anyone with access to PEQ should learn the directivity qualities they enjoy in their space and buy speakers off directivity rather than FR (as long as the speaker seems relatively easy to EQ.). Of course, it's better to have both.
Returning to the topic of the FR10, it's why I think it's a solid speaker. The frequency response isn't that bad, and the directivity behavior is excellent.
Of course, all that goes out the window if you're making recommendations for a listener who's likely to never touch EQ, which is why I appreciate Amir's occasional dual ratings for speakers or headphones.